[TCML] Re: Solid state efficiency, was: mini Tesla coil specs

[Greg Leyh]

Hi Steve,

I'd tend to agree that low voltage silicon switched tesla coils tend to 
be less efficient than HV spark gap switched systems, for the simple 
fact that a low voltage system requires far higher currents and lower 
copper losses than a typical HV SGTC design.

Most SS systems I've seen have Zchar values below an ohm, requiring 
milliohm-level copper losses to be efficient.  The coppersmithing 
required here is usually beyond the home-depot off-the-shelf approach.

Still, with the relatively few coils that I've built, the SS coil 
outperforms the SGTC's by far, in terms of spark length/kW.  The SS twin 
prototype shown here is operating at ~7kW and easily bridging 16ft:
http://www.lightninglab.org/misc/NLL_Prototype.jpg

The Zchar is only 0.75ohm, yet in can just bridge 18ft at 7kW, or about 
2.5ft/kW.  The larger 120L50k SGTC below will bridge about 25ft at 25kW, 
yielding ~1ft/kW:
http://www.lightninglab.org/gallery/2008Teslathon/images/120L02.jpg

The SS coil required a significant amount of coppersmithing to get the 
efficiency up.  But I think the perfect quenching that a SS coil offers 
may the biggest reason it outperforms the SGTC coil.  GL


Steve Ward wrote:

> Dex, Id say yes, they (low voltage silicon switched tesla coils) are likely
> to be less efficient (than HV spark gap switched systems).
> 
> In fact, the question of efficiency (in terms of spark length per input
> watt) has yielded un-clear results in all of my ventures.  Basically, the
> measurement accuracy of both power and spark length have enough error to
> cover any claim that silicon based designs are more efficient than spark gap
> designs.
> 
> My work has led me to believe a few things... A good SGTC and a good DRSSTC
> will have about the same spark length efficiency per input power.  Ive heard
> claims of SGTC performance that rival my best DRSSTC efforts (i think
> Nemesis was one of these systems that seems particularly efficient).  In any
> case, the DRSSTC doesn't seem to be exceptionally more efficient than a good
> SGTC.  What i have noticed is that a shorter driving pulse to the resonator
> (aka, energy transfer time) seems to be more efficient at producing the same
> spark lengths with less power on my large DRSSTC system.  In order to do
> this i had to lower the primary characteristic impedance so that the primary
> current/voltage would ramp up faster.  In this test the improvement was
> definitely clear, about 20% less input power for the same spark length
> performance.  But what that 20% really means to a tesla coiler, i dont
> know... The point i wanted to get to was that a SGTC system still dumps its
> bang energy into the spark *faster* than i could manage with my DRSSTC.
> Now, a SISG or OLTC (which function as the spark gap) should be able to
> achieve exactly the same energy transfer time as a SGTC, but i have yet to
> see results of one of these systems that shows it outperforming a well
> designed SGTC.  So i think energy transfer time to the spark is very
> important.  Take the extreme example of the quasi CW systems like the old
> SSTCs and VTTC system that worked off of many mS long pulses.  These systems
> can use huge "bang" energies, yet produce relatively short sparks. If the
> same bang energy was delivered in just 10's of uS, the sparks would be many
> times longer.
> 
> I wonder what ever happened to Mr Burwell... It sounds like he had a good
> grasp on this (power semiconductor) stuff, someone i wouldnt mind chatting
> with.
> 
> Steve
> 
> On Mon, Nov 9, 2009 at 5:33 AM, Dex Dexter <dexterlabs at xxxxxxxxxxx> wrote:
> 
>> This could be of some help to You guys:
>>
>>  http://scopeboy.com/tesla/t4spec.html
>>
>> I am just wondering if IGBT "gap" + primary of a very low impedance are
>> more loosy than in normal spark gaps coils at typical Tesla coil
>> frequencies.
>>
>>
>>
>> Dex